Selective N-7 alkylation of 3-methylhypoxanthine; the first synthesis of malonganenone J

Article abstract of DOI:10.1016/j.tetlet.2016.09.078

3-Methylhypoxanthine has been reacted with various alkyl halides under basic conditions. Allylic and benzylic halides reacted mostly at N-7 to give dialkylated purines, however, ring-opened imidazole by-products, probably resulting from N-1 alkylation and

Full text of DOI:10.1016/j.tetlet.2016.09.078

Tetrahedron Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Selective N-7 alkylation of 3-methylhypoxanthine; the first synthesis
of malonganenone J
⇑
Elahe Jafari Chamgordani, Jan Paulsen, Lise-Lotte Gundersen
Department of Chemistry, University of Oslo, PO Box 1033, Blindern, N-0315 Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history:
3-Methylhypoxanthine has been reacted with various alkyl halides under basic conditions. Allylic and
benzylic halides reacted mostly at N-7 to give dialkylated purines, however, ring-opened imidazole
by-products, probably resulting from N-1 alkylation and hydrolysis of the formed salt, were often
obtained in minor amounts. Less reactive halides required a larger excess, higher temperatures, and
longer reaction times. 3,7-Dialkylpurines were again the major products and imidazoles were formed
in very minor amounts, however, a considerable amount of O-alkylation also took place. The marine
natural product malonganenone J was synthesized for the first time by selective N-7 alkylation of
3-methylhypoxanthine with geranylgeranyl bromide.
Received 15 August 2016
Revised 16 September 2016
Accepted 23 September 2016
Available online xxxx
Keywords:
Alkylation
Heterocycle
Hypoxanthine
Malonganenone
Natural product
Ó 2016 Elsevier Ltd. All rights reserved.
Malonganenones are marine natural products isolated from the
Gorgonians (sea whips or sea fans) Leptogorgia sp., Echinogorgia sp.,
Euplexaura sp., and Melitodes sp.1 Malonganenone A, D, E, I, J, L, M,
and N are 7-alkylated 3-methylhypoxanthines (Fig. 1). Most of
these compounds are active against various cancer cell lines.1
Malonganenone D inhibits a receptor tyrosine kinase1c while
malonganenone L inhibits a phosphodiesterase.1d Furthermore,
malonganenone A was shown to inhibit plasmodial heat shock
protein chaperones2 and may thus be a lead compound for anti-
malarial drugs.
No total synthesis has been described for any malonganenone
to date. As a continuation of our synthetic studies directed toward
marine purine–terpene hybrid natural products and analogs,3,4 we
now report the first synthesis of malonganenone J.
We envisaged that all of the malonganenones shown in Figure 1
should be available by the N-7 alkylation of 3-methylhypoxanthine
with an appropriate allyl halide. However, there are surprisingly
few reports regarding the alkylation reactions of 3-substituted
hypoxanthines. Benzylation of compound 1a under neutral condi-
tions was reported to take place at N-1 giving salt 2, which rear-
ranged to the dibenzylated isomers 3 and 4 in almost equal
amounts (Scheme 1).5
yield an unidentified product (assumed to be compound 7, 8 or
9) and the desired 3,7-dimethylated product 6 which was only
isolated in 36% yield.6b Thus, we decided to study the
N-alkylation of hypoxanthine 1b in more detail.
3-Methylhypoxanthine (1b) is available by several literature
procedures and we chose to prepare this starting material in two
steps from adenine.7,8 First, we examined the reactivity toward
geranyl bromide (10);9 a readily available allyl halide with struc-
tural resemblance to the halides required for the syntheses of the
malonganenones (Scheme 2, Table 1). When the previously
reported reaction conditions for benzylation of compound 1b were
employed,5 only moderate conversion (45%) was observed (Table 1,
entry 1). The starting material to product ratio was significantly
improved by increasing the concentration of the reaction mixture
(Table 1, entry 2), but ring opened by-products 12a10 and 13a were
also isolated.11 Although the amount of compound 12a formed
could be estimated from the 1H NMR spectrum of the crude prod-
ucts (Table 1), similar estimates for imidazole 13a were difficult
due to overlapping signals with other compounds. The reaction
also took place at room temperature (Table 1, entry 3), but a some-
what longer reaction time was required and additional dialkylation
leading to the imidazole 12a occurred.
In the presence of a base, purines 1 have been benzylated at N-7
to give the products 5 in relatively good yields.6 Methylation of
hypoxanthine 1b, on the other hand, resulted in ring opening to
To our surprise, formation of tetraalkylammonium salt 14
(Scheme 3) was observed in the alkylations performed in DMA
(Table 1, entries 1–3). This made isolation of the products more
complicated and was probably the reason why the yields from
reactions performed under seemingly identical conditions were
not necessarily reproducible. It has been previously reported that
⇑
Corresponding author.
E-mail address: l.l.gundersen@kjemi.uio.no (L.-L. Gundersen).
http://dx.doi.org/10.1016/j.tetlet.2016.09.078
0040-4039/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Chamgordani, E. J.; et al. Tetrahedron Lett. (2016), http://dx.doi.org/10.1016/j.tetlet.2016.09.078

2
E. J. Chamgordani et al. / Tetrahedron Letters xxx (2016) xxx–xxx
Scheme 3. (a) K2CO3, rt, 1 h.
Figure 1. Structure of malonganenones A, D, E, I, J, L, M, and N.
Scheme 4. (a) K2CO3 (1.0 equiv), 15 (1.2 equiv) DMSO, rt; (b) K2CO3 (1.0 equiv), 15e
(0.75 equiv), DMSO, 12 °C.16
quaternary ammonium halides can be synthesized from tertiary
amides and alkyl halides in the presence of sodium or potassium
carbonate at elevated temperatures.12 According to the litera-
ture,12a these reactions require a temperature of at least 40 °C
and are more efficient using sodium or potassium carbonates
instead of cesium, lithium, and calcium carbonates. However,
when hypoxanthine 1b was alkylated with geranyl bromide at
ambient temperature, the ammonium salt 14 was still formed.
Compound 14 was also isolated in 22% yield when excess DMA
was reacted with geranyl bromide in the presence of K2CO3 at
ambient temperature for 1 h (Scheme 3). Changing the base to
Cs2CO3 (data not shown) or to NaH did not improve the reaction
outcome and salt 14 was also formed in DMF (Table 1, entries 4
and 5).
Scheme 1. (a)5 1a, PhCH2Br, DMA, 110 °C; (b)6 PhCH2X, K2CO3, 90–110 °C, DMA;
(c)6b (1) 1b, K2CO3, DMA 110 °C, (2) MeI, rt.
Thus, we chose to switch to a non-amide containing solvent.
When hypoxanthine 1b was alkylated in DMSO (Table 1, entries
6 and 7), full conversion and minimal ring opening took place
and the desired geranylpurine 11a was isolated in high yields.
The reactions in DMSO turned out to be more reproducible and
could be performed at ambient temperatures.
We then extended the alkylation study of 3-methylhypoxan-
thine (1b) to include benzyl (Scheme 4) and alkyl halides
Scheme 2. (a) See Table 1.
Table 1
N-Alkylation of 3-methylhypoxanthine (1b) with geranyl bromide (10)
Entry
Base (equiv)
10 (equiv)
Solvent
Conc. 1b (M)
Temp. (°C)
Timea (h)
Ratio 1b:11a:12ab
Yield 11ac (%)
Yield 12ac (%)
Yield 13ac (%)
e
e
e
1
2
3
4
5
6
7
K2CO3 (1.0)
K2CO3 (1.0)
K2CO3 (1.0)
NaH (1.2)
NaH (1.2)
K2CO3 (1.2)
K2CO3 (1.0)
1.2
1.2
1.2
1.2
1.2
1.5
1.2
DMA
DMA
DMA
DMA
DMF
0.032
0.25
0.25
0.25
0.25
110 (90d)
2.5
1.0
2.0
1.0
2.0
1.5
2.5
55:45:0
11:82:7
0:61:39
42:41:17
0:41:59
0:98:2
—
—
5
—
110 (90d)
69
36
12
e
rt
rt
rt
10
—
—
—
—
e
e
e
—
—
—
—
—
—
e
e
e
e
e
DMSO
DMSO
0.022
0.022
110 (90d)
rt
86
87
e
0:90:10
5
a
b
c
Reaction time after geranyl bromide addition.
From 1H NMR spectrum of the crude product, the amount of compound 13a could not be determined due to overlapping signals.
Isolated yield.
d
e
Temp. with base only (temp. after geranyl bromide addition).
Not isolated.
Please cite this article in press as: Chamgordani, E. J.; et al. Tetrahedron Lett. (2016), http://dx.doi.org/10.1016/j.tetlet.2016.09.078

E. J. Chamgordani et al. / Tetrahedron Letters xxx (2016) xxx–xxx
3
Scheme 5. (a) RX, K2CO3, DMSO, see Table 2.
Scheme 6. (a) (1) K2CO3, DMSO, rt, 1 h, (2) geranylgeranyl bromide, 3 h.
(Scheme 5, Table 2). Under the optimized conditions (Table 1, entry
7) benzyl bromide (15a) gave the desired product 5b in an excel-
lent yield and only ca. 2% of the benzyl analog of imidazole 12a
was observed in the 1H NMR spectrum of the crude product. Halo-
genated benzyl halides 15b–15d also gave the 7-benzylated
hypoxanthines 11b–11d in very good yields. The reaction with
1.2 equiv of the more reactive halide 15e resulted only in the for-
mation of dibenzylated ring opened product 12e (60% based on
1b, quantitative based on 15e). However, the desired product 11e
was synthesized in high yield when the amount of benzyl halide
and the reaction temperature were reduced.
Hypoxanthine 1b reacted readily with n-pentyl iodide to give
the 7-alkylated compound 11f in high yield, together with minor
amounts of ring-opened products 12f and 13f (Table 2, entry 1),
and almost identical results were obtained with n-pentyl bromide
(Table 2, entry 2). The more sterically hindered cyclohexylmethyl
bromide or iodide13 reacted slowly at ambient temperature. Only
small amounts of the desired compound 11g were formed under
standard conditions and no ring-opened products 12g or 13g were
observed. Interestingly O-alkylation leading to compound 16g also
took place (Table 2, entries 3–5).14 When 2 equiv of the bromide
were used and the reaction temperature raised to 50 °C, the con-
version was greatly improved, but both the N-7 and O-alkylated
isomers were formed (Table 2, entry 6). The less hindered cyclo-
hexylethyl bromide reacted at ambient temperature and fairly
good N-7 selectivity was observed (Table 2, entry 7). When cyclo-
pentyl bromide was used as the alkylating agent, 2 equiv of the
bromide, a reaction temperature of 50 °C, and a prolonged reaction
time was required for full conversion. In this case O-alkylation was
more efficient than the reaction with less reactive halides (Table 2,
entries 8–9). There are few prior literature reports on the synthesis
of 6-alkoxy-3-alkylpurines,15 and such compounds have not been
made by O-alkylation of a 3-alkylhypoxanthine.
Finally, the natural product malonganenone J (11j) was synthe-
sized for the first time using the developed mild and selective alky-
lation conditions (Scheme 6). When 1.2 equiv of geranylgeranyl
bromide9 were used, only 75% conversion was achieved and
malonganenone J (19) was isolated in 68% yield. When the amount
of allylation agent was increased to 1.6 equiv, 79% of the target
compound 11j was isolated and also ring opened compounds 12j
and 13j were obtained.
In conclusion, we have shown that 3-methylhypoxanthine can
be alkylated predominately at N-7 upon reaction with alkyl halides
in the presence of a base. Minor amounts of imidazoles, formed by
N-1 alkylation and subsequent ring opening of the intermediate
purinium salt, were often observed. The formation of imidazoles
12 and 13, strongly indicate that the by-product in the literature
reaction6b shown in Scheme 1, was compound 7. A substantial
amount of O-alkylation also took place when less reactive alkyl
halides were used. DMSO was shown to be a superior solvent in
these reactions compared to DMA and DMF. The marine natural
product malonganenone J was synthesized for the first time by
the reaction between 3-methylhypoxanthine and geranylgeranyl
bromide.
Table 2
N-Alkylation of 3-methylhypoxanthine (1b) with alkyl halides
Entry
K2CO3 (equiv)
RX (equiv)
Temp. (°C)
Timea (h)
Unconverted 1bb
Yield 11c (%)
Yield 16c (%)
Yield 12c (%)
6, 12fd
Yield 13c (%)
<5, 13f
e
1
2
1.0
1.0
CH3(CH2)4I (1.2)
CH3(CH2)4Br (1.2)
rt
rt
3.0
3.8
None
None
84, 11f
89, 11f
—
e
—
7, 12f and 13f 1:4 ratio
e
e
e
e
e
e
e
f
3
4
5
6
7
1.0
1.2
1.2
1.2
1.0
rt
4.0
24
24
24
4.0
ca. 85%
ca. 65%
ca. 60%
ca. 17%
None
15, 11g and 16g 20:1 ratio
—
—
—
—
—
—
—
—
—
(1.2)
(1.6)
(1.6)
(2.0)
rt
30, 11g and 16g 19:1 ratio
f
rt
25, 11g
56, 11g
65, 11h
—
f
rt (50g)
rt
14, 16g
6, 16h
5, 13h
(1.2)
(1.2)
e
e
e
8
9
1.0
1.2
rt
4.0
30
ca. 80%
None
11, 11i
48, 11i
6, 16i
—
—
—
rt (50g)
23, 16i
5, 13i
(2.0)
a
b
c
d
e
f
Reaction time after RX addition.
From 1H NMR spectrum of the crude product.
Isolated yield.
Contained 10% of the isomer resulting from 1,9 dialkylation of the starting hypoxanthine followed by ring opening.
Not observed.
Formed in minor amounts but not isolated in pure form.
Temp. with base only (temp. after RX addition).
g
Please cite this article in press as: Chamgordani, E. J.; et al. Tetrahedron Lett. (2016), http://dx.doi.org/10.1016/j.tetlet.2016.09.078

4
E. J. Chamgordani et al. / Tetrahedron Letters xxx (2016) xxx–xxx
Charnock, C.; Felth, J.; Gørbitz, C. H.; Larsson, R.; Tamm, T.; Gundersen, L.-L.
Acknowledgments
Pure Appl. Chem. 2011, 83, 645–653; (m) Aarhus, T. I.; Fritze, U. F.; Hennum, M.;
Gundersen, L.-L. Tetrahedron Lett. 2014, 55, 5748–5750.
The Norwegian Research Council is gratefully acknowledged for
financial support (Grant No. 209330) and for partly financing the
NMR instruments used.
5. Montgomery, J. A.; Hewson, K.; Thomas, H. J. J. Org. Chem. 1966, 31, 2202–2210.
6. (a) Montgomery, J. A.; Thomas, H. J. J. Org. Chem. 1963, 28, 2304–2310; (b) Fujii,
T.; Saito, T.; Inoue, I.; Kumazawa, Y.; Tamura, K. Chem. Pharm. Bull. 1988, 36,
107–117.
7. Adenine was methylated at N-3 as described in Ref. 3e, and reacted with HNO2
to give the target compound 1b according to: Itaya, T.; Matsumoto, H. Chem.
Pharm. Bull. 1985, 33, 2213–2219.
Supplementary data
8. Synthesis of compound 1b according to: Wright, A. E.; Roth, G. P.; Hoffman, J.
K.; Divlianska, D. B.; Pechter, D.; Sennett, S. H.; Guzman, E. A.; Linley, P.;
McCarthy, P. J.; Pitts, T. P.; Pomponi, S. A.; Reed, J. K. J. Nat. Prod. 2009, 72, 1178–
1183. gave, in our hands, not completely pure product and unidentified by-
products(s) which interfered with the alkylation reactions.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2016.09.
078.
9. E-Geranyl bromide and geranylgeranyl bromide were synthesized from E-
geraniol or geranylgeraniol respectively, as previously described.4b
10. A similar product has been formed from 1,3,7-tribenzylated hypoxanthine;5
see also the proposed structure 7 in Scheme 1.6b
11. Structure elucidation was based on NMR spectroscopy, especially HMBC and
HMQC. For all compounds 11 it was determined that alkylation had taken place
at N-7 based on the 3 bond correlation between the protons in the CH2 or CH
attached to N-7 and C-5 (the quartenary carbon in the purine with an HMBC
correlation to H-8 but not to H-2) and C-8 (the CH in the purine not correlating
with the methyl group at N-3) in the purine.
References and notes
1. (a) Keyzers, R. A.; Gray, C. A.; Schleyer, M. H.; Whibley, C. E.; Hendricks, D. T.;
Davies-Coleman, M. T. Tetrahedron 2006, 62, 2200–2206; (b) Sorek, H.; Rudi, A.;
Benayahu, Y.; Ben-Califa, N.; Neumann, D.; Kashman, Y. J. Nat. Prod. 2007, 70,
1104–1109; (c) Zhang, J.-R.; Li, P.-L.; Tang, X.-L.; Qi, X.; Li, G.-Q. Chem. Biodivers.
2012, 9, 2218–2224; (d) Sun, Z.-H.; Cai, Y.-H.; Fan, C.-Q.; Tang, G.-H.; Luo, H.-B.;
Yin, S. Mar. Drugs 2014, 12, 672–681.
2. (a) Cockburn, I. L.; Pesce, E.-R.; Pryzborski, J. M.; Davies-Coleman, M. T.; Clark,
P. G. K.; Keyzers, R. A.; Stephens, L. L.; Blatch, G. L. Biol. Chem. 2011, 392, 431–
438; (b) Cockburn, I. L.; Boshoff, A.; Pesce, E.-R.; Blatch, G. L. Biol. Chem. 2014,
395, 1353–1362.
12. (a) Ropponen, J.; Lahtinen, M.; Busi, S.; Nissinen, M.; Kolehmainen, E.; Rissanen,
K. New J. Chem. 2004, 28, 1426–1430; (b) Busi, S.; Lahtinen, M.; Ropponen, J.;
Valkonen, J.; Rissanen, K. J. Solid State Chem. 2004, 177, 3757–3767; (c) Kärnä,
M.; Lahtinen, M.; Valkonen, J. J. Mol. Struct. 2009, 922, 64–76.
3. For a recent review of marine purine-terpene hybrid natural products, see:
Gundersen, L.-L. Phytochem. Rev. 2013, 12, 467–486.
13. Cyclohexylmethyl iodide was synthesized from cyclohexylmethanol in 89%
yield following the general procedure for conversion of primary alcohols to
iodides found in: Liu, W.; Li, L.; Li, C.-J. Nat. Commun. 2015, 6, 6524.
14. The structure elucidation of the O-alkylated isomers 16 was based on the
following: In HMBC, the OCH or OCH2 protons correlated with only one carbon
in the purine identified as C-6 (the quartenary carbon with a correlation to H-2
but not to H-8). 15N HMBC revealed no correlation between the OCH or OCH2
protons and any of the nitrogen atoms. In 1D selective NOESY a strong NOE
effect was seen between the CH3 protons and H-2 and a weak effect between
H-2 and the OCH or OCH2 protons.
15. (a) Townsend, L. B.; Robins, R. K. J. Am. Chem. Soc. 1962, 84, 3008–3012; (b)
Bergmann, F.; Neiman, Z.; Kleiner, M. J. Chem. Soc. C 1966, 10–15; (c) Diller, D.;
Neiman, Zohar; Bergmann, F. J. Chem. Soc. C 1968, 878–881; (d) Barlin, G. B.;
Fenn, M. D. Aust. J. Chem. 1983, 36, 633–688; (e) Kohda, K.; Baba, K.; Kawazoe,
Y. Tetrahedron Lett. 1987, 28, 6285–6288.
4. (a) Utenova, B. T.; Gundersen, L.-L. Tetrahedron Lett. 2004, 45, 4233–4235; (b)
Bakkestuen, A. K.; Gundersen, L.-L.; Petersen, D.; Utenova, B. T.; Vik, A. Org.
Biomol. Chem. 2005, 3, 1025–1033; (c) Vik, A.; Hedner, E.; Charnock, C.;
Samuelsen, Ø.; Larsson, R.; Gundersen, L.-L.; Bohlin, L. J. Nat. Prod. 2006, 69,
381–386; (d) Vik, A.; Gundersen, L.-L. Tetrahedron Lett. 2007, 48, 1931–1934;
(e) Vik, A.; Hedner, E.; Charnock, C.; Tangen, L. W.; Samuelsen, Ø.; Larsson, R.;
Bohlin, L.; Gundersen, L.-L. Bioorg. Med. Chem. 2007, 15, 4016–4037; (f) Roggen,
H.; Gundersen, L.-L. Eur. J. Org. Chem. 2008, 5099–5106; (g) Sjögren, M.;
Dahlström, M.; Hedner, E.; Jonsson, P. R.; Vik, A.; Gundersen, L.-L.; Bohlin, L.
Biofouling 2008, 24, 251–258; (h) Proszenyák, Á.; Charnock, C.; Hedner, E.;
Larsson, R.; Bohlin, L.; Gundersen, L.-L. Arch. Pharm. Chem. Life Sci. 2007, 340,
625–634; (i) Proszenyák, Á.; Brændvang, M.; Charnock, C.; Gundersen, L.-L.
Tetrahedron 2009, 65, 194–199; (j) Vik, A.; Proszenyák, Á.; Vermeersch, M.; Cos,
P.; Maes, L.; Gundersen, L.-L. Molecules 2009, 14, 279–288; (k) Roggen, H.;
Charnock, C.; Burman, R.; Felth, J.; Larsson, R.; Bohlin, L.; Gundersen, L.-L. Arch.
Pharm. Chem. Life Sci. 2011, 344, 50–55; (l) Roggen, H.; Bohlin, L.; Burman, R.;
16. The same set of reaction conditions as shown in Table 1, entry 7 was used
unless otherwise specified.
Please cite this article in press as: Chamgordani, E. J.; et al. Tetrahedron Lett. (2016), http://dx.doi.org/10.1016/j.tetlet.2016.09.078